The effect of elevated mixed venous oxygen tension(PvO2) on the diffuse alveolar hypoxia was studied in dogs using venovenous(VV) extracorporeal lung assist(ECLA). Six mongrel dogs were mechanically ventilated with the continous infusion of pentobarbital. A double lumen tube was inserted via the right external jugular vein and was eonnected with the ECLA cireuit to establish a VV bypass. A Kurare oxygenator 0.3m2 was chosen to obviate the use of homologous blood for priming. The total volume of the ECLA circuit was 150mL Without ventilating the oxygenator during VV ECLA, we decreased F1O2 from 0.21 to 0.1 via 0.15 to evaluate the hypoxic repsonse of lung. Stepwise reductions in F1O2 0.21 to 0.l caused the arterial oxygen tension(PvO2) and (PvO2 to decrease while the mean pulmonary arterial pressure(MPAP) and pulmonary vascular resistance(PVR) progressively increased. We hypothesized that the reduction of F1O2 without aceompanying decrease of PvOmight not induce hypoxic pulmonary vasoconstriciton(HPV) which was proved at low F1O2 with low PvO VV ECLA was tried on another 7 dogs while monitoring arterial oxygen saturation(SaO2) and mixed venous oxygen saturation(SvO2) by two oximetrix catheters. The elevation of SvO2 during VV ECLA was followed by the elevation of SaO2 We compared MPAP and PVR at high F1O2 with those at low F1O2with VV ECLA while making SaO2 equal. PvO2 were 39±11mmHg at F1O2 0.21 and 62±11mmHg at F1O2 0.15 with VV ECLA. PvO2 were 30±8mmHg at F1O2 0.15 and 53±10mmHg at F1O2 0.1 with VV ECLA. MPAP and PVR were 18±5mmHg and 176±56 dyne sec/cm5 at F1O2 0.21 and 19±4mmHg and 198±94 dyne sec/cm5 at F1O2 0.15 with VV ECLA . MPAP and PVR were 25 5 mmHg and 430±250 dyne. sec/ cm5 at F1O2 0.15 and 25±5mmHg and 400±197 dyne sec/cm5 at F1O2 0.1 with VV ECLA. Decrease of F1O2 from 0.21 to 0.15 and from 0.15 to 0.1 did not cause significant ehanges in MPAP and PVR during VV ECLA. Our findings indicate that small increase of PvO2 by VV ECLA may prevent or diminish hypoxic resyonse of the whole lung.